DESCRIPTION: (Applicant's Description) This project aims to generate cDNA libraries from fresh pediatric leukemia cells that have been directly isolated from blood at the time of diagnosis. Cases that represent the major subtypes of childhood Acute Lymphoid leukemia (ALL) and at least one subtype of Acute Myelogenous Leukemia (AML) will be chosen for analysis. The library construction methodology will utilize improvements on existing techniques which have been demonstrated to generate libraries with a high proportion of foil-length clones. Clones from these libraries will contain full-length copies of the original mRNA, and the libraries will be representative of the original messenger population. The libraries will be normalized to reduce the variation in abundance of different clones and aliquots from non-normalized and normalized libraries will be arrayed to facilitate replication and identification of clones of interest. EST sequencing on 10,000 randomly selected clones will be performed each year to select cDNAs that appear to be full- length, and for which there is no existing foil length sequence available in the databases. The full length sequence of >350 1.5-2.0 kb cDNAs will be determined each year, using a concatenation cDNA sequencing approach that was developed in this laboratory. This protocol enables us to simultaneously perform full-length sequencing on as many as 70 cDNAs after construction of a single shotgun library. A PCR-based method that has been optimized m our laboratory will be used to identify the 5'-ends of clones that are not full length. All data will be released immediately and all clones will be freely available. These malignant tissues have not been extensively sequenced, and can therefore be expected to provide substantial new information compared to that derived from leukemia cell lines. Hence, the availability of these libraries will provide a mechanism for rapidly isolating full-length copies of previously identified partial clones, and will provide a source for identifying new cancer-related genes. The integration of the tissue collection, library construction, EST analysis, full length clone insert sequencing and 5'- end rescue will define a complete and scalable pathway for the systematic analysis of genes involved in cancer.